The stabilization of man-made artificial systems has been achieved by sensor based state feedback control with high computational bandwidth and high stiffness structures. In contrast, many biological systems have been achieved similar or superior stable behavior with low speed signal transmission via nervous systems, which is easy to introduce unstable performance from a control engineering perspective. In order to explain this phenomenon, the concept of self-stabilization has recently been proposed and investigated. Self-stabilization is defined as the ability to restore its original state after a disturbance without any feedback control. In this paper, the self-stabilizing function of a musculoskeletal system for arbitrary motion in the vertical plane is analytically investigated using Lyapunov stability theory. Based on this investigation we propose a design method to realize the self-stabilizing function of a musculoskeletal system, and experimentally verify that the self-stabilizing function can be physically realized by the proposed Lyapunov function.